The overall aim of this project is to test our hypothesis that the intrinsic communication of neural systems in the absence of any external goals or task demands (i.e., in a resting state), predicts individual and age differences in implict probabilistic sequence learning (IL). IL is a type of learning that occurs without conscious awareness or intent and is central to many fundamental life skills, such as adapting to ever-changing technologies and environments. It is important to examine the neural mechanisms underlying the age-related deficits in IL that are typically observed when older adults are compared to younger ones because being able to learn in this way can have implications for maintaining independence and quality of life. How brain regions communicate with each other in a resting state--when mental activity is unconstrained--has recently emerged as an important neural mechanism underlying individual and age differences in cognitive functioning. A measure of resting state communication, called resting state functional connectivity (rsFC) has even been found to predict future individual differences in cognitive performance, suggesting that rsFC may be promoting a neural system's readiness to perform cognitive tasks. So far, however, research on the predictive value of rsFC for performance on specific cognitive processes, such as learning, is limited. The proposed study will be the first to examine whether rsFC can predict individual and adult age differences in a type of implicit learning in which people must become sensitive to subtle probabilistic patterns. Aim 1 will measure rsFC before healthy young and older adults complete a measure of IL, the Triplets Learning Task (TLT), and will assess whether individual differences in rsFC between task-relevant regions predict IL in the TLT. It is hypothesized that more positive rsFC between the caudate and MTL, two brain regions known to be coactivated during IL, will predict better IL in both young and older adults and also mediate age differences in IL. These hypotheses are based on the preliminary study for this proposal, which found the predicted relationship between caudate-MTL connectivity and IL in younger adults, and on previous fMRI studies showing that age differences in IL are related to the task-evoked activity of these regions. Aim 2 will assess how rsFC, measured prior to the TLT, relates to patterns of connectivity during the TLT, and how changes in connectivity between the rest and task states relate to IL. We hypothesize that (1) young adults will show greater changes in connectivity from rest to task than older adults and (2) Greater changes in connectivity between the caudate and MTL from rest to task will be related to better IL for both age groups. These hypotheses are based on studies showing that functional connections between brain regions change based on cognitive state, and on studies showing declines in the state-related adaptability of functional connections with age. Characterizing the neural systems underlying IL will foster a better understanding of the changes in essential behavioral functions that occur in healthy aging.